But not all the problems are resolved, eg:

• The MIGPOP adjustment for the 1991 undercount is now seen as too large (after the 2001 Census results)
• Not all the migrant students in 2001 should be deducted, as some will have moved home with parents (deemed a migration already in 1991, even if parents moved while child was away at university!)
• The geography of recorded migration by recent graduates in 2001 (now from termtime address) will differ from that in 1991 (then from parental address) NB: wouldn’t even have been recorded as a migration in 1991 if they were back with parents then
• No allowance for differences between ‘origin not stated’ in 1991, and imputation plus ‘no usual address one year ago’ in 2001 (re. net migration by district)